From the early beginnings of computer science, researchers have tried to build systems that are more capable than humans in specific tasks. In particular, tasks have been addressed that are usually associated with intelligence. These have often been games. For instance, computer systems are now the best players in Chess, Jeopardy, Go, and Poker. However, decades of research in artificial intelligence have also revealed that many everyday tasks are much harder to master for artificial systems than tasks that are hard for humans. Therefore, the RoboCup Federation has set itself the goal that "by the mid-21st century, a team of fully autonomous humanoid soccer players shall win the soccer game against the winner of the most recent World Cup". This is very ambitious, because soccer players do not only need a mind, they also need a body, and they act as a team. The problem is addressed with several different soccer leagues, some just simulated, some with driving robots, and some with humanoids. The talk will give an insight in how such systems work with the 2016 Standard Platform League world champion team B-Human as an example.
Thomas Röfer received his Diploma degree and PhD degree in computer science from the University of Bremen in 1993 and 1998. He was member of the DFG Transregional Collaborative Research Center “Spatial Cognition: Reasoning, Action, Interaction” between 2003 and 2014. Since 2006, he has been with the German Research Center for Artificial Intelligence (DFKI) in the department for Cyber-Physical Systems. He has been active in RoboCup since 2001, first as the speaker of the three-times world-champion “GermanTeam” in the Four-legged League, then as the leader of the five-times world champion team “B-Human” in the Standard Platform League. His research interests include humanoid robots, real-time computer vision, robot simulation, and behavior control.